Drilling for oil is not easy; you already know that. It's a harsh and unforgiving operation even under benign circumstances, and unimaginably worse when the climate is bitterly cold, or the platform is in deep water with high waves and storms as a regular occurrence. Drilling even a “simple” hole straight down brings enormous instrumentation challenges, beginning with the obvious: How do you know the drilled hole really is perfectly vertical? Even a tiny deviation or drift in the angle means you could be way off when you are go down a few thousand feet or deeper. Differential GPS with its centimeter-range accuracy, or ground-penetrating radar are not options here, of course.
The basic instrumentation incurs downhole temperatures of 200°C or more — not an easy place to get reliable electronics or consistent data, especially when you are looking for accuracy to a fraction of a degree. Techniques previously used included dropping a weighted line to punch a target disk at the drill tip site, and using a glass container filled with sulfuric acid to etch a horizontal axis. These techniques and others worked to a certain point, but required costly, time-consuming halts to the drilling process and were not compatible with the desire for constant feedback and corrective action.
Now add to this challenge the fact that many new drilling sites aren't just those basic vertical holes, but go horizontally as well, often with multiple downholes and their branches, all starting at a single top-side pad. How you get a drilling bit and its pipe to make such a right-angle turn at all, let alone so far down — that's an amazing story itself, but not the issue here. How do you make sure each downhole is true and, even more difficult, that the horizontal boring is also on track, where you have to worry about angular errors and deviations in both horizontal and vertical extensions?
Fortunately, you can read a well-written and very illuminating piece about previous approaches as well the most advanced ones in the Autumn 2013 issue of Oilfield Review (thanks to Rick DeMeis for bringing this article to my attention). The lengthy article Geomagnetic Referencing — The Real-Time Compass for Directional Drillers is well worth reading: It describes how advanced accelerometers, gyroscopes, and magnetometers, combined with near-real-time three-dimensional geomagnetic mapping at multiple layers of the Earth's ever-changing magnetic field, are used to provide highly accurate results.
It's an amazing combination of drill-head sensors and electronics, site and pad instrumentation, local geomagnetic measurement, and extremely sophisticated data-analysis and predictive algorithms, and it is all fully field-tested.
Just as impressive as the results themselves is that all this is accomplished while drilling without the need to stop; appropriately, it's called Measurement While Drilling. Thus, the new approach of integrating sensors and local-area information with calculation- and algorithm-intensive processing has solved three problems: the downhole vertical problem, the horizontal drilling problem, and the halt-while-measuring problem.
Too often, people who casually talk as if they are knowledgeable about advanced technologies either downplay or are just ignorant about what's really involved. In short, “what's the big deal?” is the attitude. (Believe me, I also have the highest regard for those crew who can remotely pour concrete at the sea floor, 2000+ feet underwater!)
It seems to me that a large part if the problem is not only their willful ignorance, but is also an unintended consequence of our success. The fact that the scientists, engineers, technicians and on-site workers routinely perform what are, frankly, near-miraculous missions on a regular basis makes it all look routine and easy, when they are anything but that.
I view this “extreme design” as one of the least-appreciated, least-discussed aspects of engineering and technology, though you will sometimes see them in special TV shows on niche channels. If these types of stories interest you — and I hope they do — check out the list here of such stories which EE Times and EDN have done in the past years. They make clear that much of what I'll broadly call “engineering” is pretty hard in practice, even if it is straightforward in concept.
Just think about the Voyager 1 (and soon, Voyager 2) space probe, and the incredibly minuscule power levels and accounted SNR of the signals coming back from it as it leaves the outer boundaries of our solar system (see NASA link here as well as the book Voyager: Seeking Newer Worlds in the Third Great Age of Discovery by Stephen J. Pyne). And how do they drill those very long rail/auto tunnels through mountains, starting at both ends, and meet perfectly midway?
Have you ever been involved with an especially “extreme” test and measurement or analog-signal problem? Did you and your team anticipate it reasonably well, or not?